Abstract
Human reaction times (RTs) for finding a target unique in orientation (O) among homogeneous non-targets are shortened when this target is also unique in color (C). Let RTC, RTO, or RTCO be the RT for a target unique in C, O, or in both C and O. RTCO is typically shorter than predicted by a race model RTCO=pmin(RTC,RTO) (here, min(x,y) denotes the minimum of x and y, and x=py means x and y have the same probability distribution), characterizing the statistical facilitation between C and O [1-3]. According to the V1 Saliency Hypothesis (V1SH), this race model violation implicates a contribution to saliency by CO neurons conjunctively tuned to C and O. The RT for a target unique in O is also shortened by making the target also unique in eye of origin (E), when the target is shown to a different eye from that for the non-targets. This demonstrates a saliency by E, and implicates the primary visual cortex (V1) [4] since V1 is the only cortical area with monocular neurons for encoding E. However, one could not quantify the saliency by E by directly measuring RTE for finding a target unique in E only, since E is perceptually non-distinctive or even invisible [5]. By making the target unique in E additionally, a target C, O, or CO becomes a target CE, EO, or CEO, respectively. In this study, we first measure the RTs for these six target types. Our data confirm that RTO<RTEO, and reveal that RTC<RTCE and RTCO<RTCEO, extending the previous findings. Furthermore, V1SH, which links visual saliency (and thus the search RT) with V1 neural activities, establishes [6]:
min(RTCEO,RTC,RTO,RTE)=pmin(RTCO,RTCE,RTEO).
This equation enables us to calculate RTE from our measured RTs, thus quantifying the saliency of E. Finally, we infer the contribution to saliency by any CE or EO neurons, conjunctively tuned to C and E, or to E and O, respectively. A significant contribution should also suggest an abundance of such neurons. We infer this by examining whether RTCE or RTEO is shorter than predicted by the corresponding race model. According to our data, this contribution may be stronger by the EO than the CE neurons, and depends on the colors of both the target and non-targets. Our method provides a tool to quantify the saliency by the perceptually non-distinctive E, and we demonstrate a nontrivial application of this quantification.
